PROJECT SUMMARY Applicant. This pre-doctoral fellowship (F31) application is designed to promote the training of Joyce Rhoden, a pre-doctoral student in the Department of Epidemiology at the University of North Carolina. Her goal is to become an independent researcher who conducts innovative, translational health research. During the F31 training the applicant will be mentored by her sponsors Drs. Marilie Gammon and Cathrine Hoyo. Significance. Breast cancer is the most commonly diagnosed cancer among women in the United States. Polycyclic aromatic hydrocarbons (PAHs) result from incomplete combustion and are ubiquitous. Many, but not all, studies show positive associations between PAHs and breast cancer risk. Vitamin D significantly protects against oxidative stress in breast epithelial cells and is associated with lower breast cancer risk. Vitamin D prevents pre-neoplastic lesions in mammary gland explants after exposure to the PAH carcinogen 7,12- dimethylbenz(a)-anthracene (DMBA). However, no epidemiologic study has examined if associations between PAHs and breast cancer incidence vary with plasma vitamin D or vitamin D polymorphisms. The study hypothesis is that elevated risks for breast cancer incidence associated with PAHs will be evident only in women with low vitamin D levels, or with genotypes that may lower vitamin D levels. Aim 1. Determine if PAH measures (PAH-DNA adducts and PAH sources (active and passive smoking, grilled/smoked foods, indoor fireplace use, vehicular traffic)) interact with plasma 25-hydroxyvitaminD (25-OHD) concentrations to influence breast cancer risk. Aim 2. Determine if PAH measures interact with any of 25 vitamin D-related genetic single- nucleotide polymorphisms (SNPs) to influence breast cancer incidence. Aim 3. Determine if genetically predicted vitamin D levels influences breast cancer risk and assess whether the genetically predicted vitamin D levels modify the PAH-breast cancer associations. Approach. The dissertation will use existing information from 1508 women with breast cancer and 1556 women without breast cancer. Measures of PAH and vitamin D were assessed shortly after diagnosis. For Aims 1-3, logistic regression will be conducted to estimate odds ratios and 95% confidence intervals to examine interaction on the multiplicative and additive scales. For Aim 3, an adjusted two-stage least squares regression method will be used to estimate a weighted genetic predictive score, which can then be used in logistic models. Innovation. This dissertation: (1) will be the first epidemiologic study to examine interaction between PAHs and vitamin D in association with breast cancer risk; (2) includes multi-level PAH and vitamin D measures; (3) explores biologically plausible SNPs; (4) will estimate genetically predicted vitamin D levels; and (5) uses a well-characterized population-based sample. Impact. The proposed dissertation may identify a potential low-cost risk reduction strategy for breast cancer.